Nail down table of contents

[franciscothebrady]

1. intro: what is the aim of the book
   • Why do social science practitioners need programming and statistical skills
   • etc
2. introduction to R: beginning steps
   • begin with case study of social science data project
   • introduction to data types in R
   • extracting data stored in different formats (pdfs, csv, etc)
3. statistical concepts for effective for social science research
   • descriptive statistics: what we know
   • correlation - connecting the dots
   • causal inference - establishing a direction
   • external validity - controlling for confounders (maybe move to later chapter)
4. longer projects:
   • version control: git
   • reproducable research
   • scandal!

[franciscothebrady]

very preliminary

[damianrt]

Possible Topics or sections

Carry out various types of data analysis: qualitative, quantitative

Data Structures as they relate to social science concepts

• structured: tabular, hierarchical
• Unstructured, text, images, other
• univariate (vectors)
• Multivariate (data frames, Matrices)

Statistical tests: t, Chisq, Anova

Modelling: Regression, lm, Glm, Fitting arbitrary models, Optim

The Formal empirical research process:

1. Ask a question
2. Formulate a hypothesis. Make a prediction based on assumed relationships between measurable concepts, aka theory.
3. Collect data
4. Analyze data
5. Evaluate Hypothesis and refine theory if necessary

[franciscothebrady]

Some people to shamelessly steal from:

• Urban Informatics Toolkit
• Computing for the Social Sciences - In particular the section on shapefiles.

[franciscothebrady]

Data Acquisition (Working with APIs): https://data.library.virginia.edu/using-census-data-api-with-r/

[franciscothebrady]

@damianrt just wrote the git addendum

[franciscothebrady]

https://mgimond.github.io/Spatial/reading-and-writing-spatial-data-in-r.html

[damianrt]

In addition to the data structures R uses to store data, we should also cover the various types of variables from a data analysis perspective: numerical vs categorical, continuous vs discrete, ordered and unordered.

[franciscothebrady]

I think variable types comes before data structures right ?

[damianrt]

I agree. I think it is useful to introduce vectors and the data types before introducing data frames etc. if we want to be able to introduce programming or more advanced usage later.